 |
INTRODUCTION |
Poly(ADP-ribosyl)ation is a covalent post-translational
modification of nuclear proteins that is induced by DNA strand breaks and represents an immediate cellular response to DNA lesions caused by
environmental insults (1). Extensive but transient pADPr synthesis is
triggered within the nucleus after treatment of mammalian cells with
genotoxic agents such as ionizing radiation or alkylation (2, 3).
Poly(ADP-ribosyl)ation also takes place during many processes such as
DNA repair (4-7), replication (8, 9), and transcription (10, 11),
regulation of cell cycle progression (12, 13), and cell differentiation
(1, 14).
Poly(ADP-ribose) polymerase 1 (PARP-11; EC 2.4.2.30), the
only well characterized enzyme in animal cells that catalyzes pADPr synthesis, comprises three functional domains: a DNA-binding domain, an
automodification domain, and a catalytic domain (15). The amino-terminal DNA-binding domain (apparent molecular mass 46 kDa)
contains two zinc fingers, responsible for the recognition and binding
to both single and double DNA strand breaks, as well as a bipartite
nuclear localization signal (16). The 55-kDa carboxyl-terminal domain
that bears the NAD+-binding site and the catalytic activity
represents the most highly conserved part of the enzyme (3, 17, 18).
The enzymatic activity of the COOH-terminal domain is greatly enhanced
upon the binding of PARP-1 to DNA strand breaks (15, 19-21). The
central 16-kDa automodification domain is rich in glutamic acid
residues, which are sites for covalent binding of pADPr (22). This
domain contains a BRCT domain, a putative site of interactions with
other proteins (23, 24).
To better define the biological functions of PARP-1 and
poly(ADP-ribosyl)ation, PARP-1 (
/) mice have been independently generated by three groups (25-27). The primary phenotype in these animals is genomic instability following DNA damage (26, 28). Although
a slightly reduced proliferation rate is observed in derived PARP-1
(/) cells, no effect on cell cycle profile or on cell
differentiation has been reported. Moreover, the homozygous mutant mice
develop normally and are fertile (29). These observations are hard to
reconcile with previous studies showing that poly(ADP-ribosyl)ation is
involved in many important processes related to DNA metabolism. However, recent studies have shown that PARP-1 (/) cells retain the
ability to synthesize pADPr following treatment with genotoxic agents,
suggesting the presence of an as yet unidentified enzyme with this
activity (30). This idea is supported by the discovery of two
structurally different and functionally active PARP proteins in plants
(31, 32) as well as a human PARP-like protein, called tankyrase (33).
Three recent papers describe the cloning of new PARP family members
cDNAs from mouse and human cells (PARP-2 and PARP-3)
(34-36).
In this paper, we show that a poly(ADP-ribose) polymerase activity is
present in immortalized PARP-1 (/) embryonic mouse fibroblasts
(25), and report the cloning of a cDNA that encodes a protein
responsible for pADPr synthesis. The enzyme, designated sPARP-1 (for
short poly(ADP-ribose)
polymerase 1), is identical to the catalytic domain of
PARP-1 and shares most of the well documented features of the
carboxyl-terminal part of PARP-1. The determination of the chromosomal
localization of sPARP-1 gene leads us to conclude that both
sPARP-1 and PARP-1 are products of the PARP-1 gene. sPARP-1
is present in wild type and PARP-1 (/) cells and is localized in
the nucleus. Although the poly(ADP-ribose) polymerase activity of
sPARP-1 is DNA strand break-independent, it is strongly stimulated by
genotoxic treatments such as alkylation and UV irradiation, suggesting
the involvement of PARP-1 and sPARP-1 in different types of DNA
damage-inducible response pathways.
| |
EXPERIMENTAL PROCEDURES |
Reagents--
The enzymes poly(ADP-ribose) polymerase-1 and
poly(ADP-ribose) glycohydrolase were purified as described previously
(37, 38). The PARP-1 inhibitors 3-aminobenzamide (3-AB),
1,5-dihydroxyquinolinediol, and 3-methoxybenzamide were obtained from
Sigma-Aldrich Canada (Oakville, Ontario), as well as the alkylating
agent,
N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), and DNase I-activated DNA. All reagents were of analytical grade.
Cell Culture--
The mouse fibroblast L (PARP-1 (+/+)), the
PARP-1 (/) mouse embryo fibroblast A1, and the A1 mother cell line
F20 (PARP-1 (+/+)) (25) were grown at 37 °C in a humidified 5%
CO2 atmosphere. The medium (Life Technologies, Inc.) was
modified Eagle's medium containing 0.2% bicarbonate for L cells, and
Dulbecco's modified Eagle's medium low glucose supplemented with 1%
L-glutamine and 0.2% bicarbonate for A1 and F20 cells.
Penicillin (100 µg/ml), streptomycin (100 µg/ml), and 10% fetal
bovine serum were added to both media. A1 cells were always maintained
in presence of 600 µg/ml neomycin (Life Technologies, Inc.) as a
selection agent (25).
Treatment of Cells--
Cells grown to 80% confluence were
treated at 37 °C for various times with either 100 µM
MNNG in serum-free medium or with a germicidal lamp (G. W. Gates & Co. Inc., New York) at a dose of 30 J/m2 as verified with
an ultraviolet meter (Ultra-Violet Products Inc., Upland, CA). To study
the inhibition of pADPr synthesis, cells were preincubated with 2 mM 1,5-dihydroxyquinolinediol for 5 min at 37 °C.
PADPr and NAD+ Quantification--
PADPr and
NAD+ were affinity purified on a dihydroxyboronyl-Bio-Rex
70 matrix as described earlier (39) except that the elutions were
performed with water at 37 °C. The amount of synthesized pADPr was
estimated by immunodot blot (40), while the analysis and measurements
of NAD+ were performed using an enzyme cycling assay
(39).
Cloning of sPARP-1 cDNA--
Total RNA was extracted from
PARP-1 (/) cells according to the method of Chomczynski and Sacchi
(41). Three µg of RNA were reverse-transcribed at 42 °C using
oligo(dT)16 primers (Amersham Pharmacia Biotech) and
Moloney murine leukemia virus reverse transcriptase (Promega) according
to the instructions of the manufacturer. For PCR amplification, primers
were chosen based on the cDNA sequences present in the catalytic
domain of PARP-1 that were most conserved among species (5'-primer:
5'-CTTCTGGAGGACGACAAGGA-3'; 3'-primer: 5'-CCACAGGGATGTCTTAAAAT-3').
Restriction sites EcoRI and HindIII were added to
the 5'- and 3'-primers, respectively. Five µl of the reverse
transcription reaction mixture was directly added to 95 µl of PCR
reaction mixture containing thermopol buffer, 200 µM of
each dNTP, 100 mM MgSO4, 15 pmol of each 3'-
and 5'-primer, and 0.8 units of Vent DNA polymerase for 30 cycles at
94 °C for 1 min, 60 °C for 1 min. and 72 °C for 2 min. We
obtained thus a partial cDNA of sPARP-1, referred to as sPARPcat,
which was cloned into EcoRI-HindIII restriction
sites of pBlueScript SK
(Stratagene).
The 5' end of the sPARP-1 cDNA was amplified using the Marathon
cDNA amplification kit from CLONTECH.
First-strand cDNA was obtained by reverse transcription of 1 µg
of total RNA extracted from A1 cells catalyzed by 100 units of Moloney
murine leukemia virus reverse transcriptase and using
oligo(dT)16 primer (Amersham Pharmacia Biotech) in a final
reaction volume of 15 µl. To synthesize the second-strand, 10 µl of
the reverse transcription reaction mixture was incubated with 40 units
of Escherichia coli DNA polymerase I, 1.5 units of RNase H,
and 15 units of E. coli DNA ligase in a final volume of 80 µl according to the CLONTECH protocol. After 1.5 h of incubation at 16 °C, 10 units of T4 DNA polymerase
were added to the reaction mixture and further incubated at 16 °C
for 45 min to create blunt ends on the double-stranded cDNA. The
Marathon cDNA adaptor was ligated to the 5' and 3' extremities of
the phenol-purified double-stranded cDNA, using 1 unit of T4 DNA
ligase to catalyze this reaction at 16 °C overnight. 5'-Rapid
amplification of cDNA ends was performed with an adaptor primer
supplied with the kit (AP2; CLONTECH) and the 3'
gene-specific primer described above (30 cycles at 94 °C for 30 s, 60 °C for 30 s, and 68 °C for 3 min). The correct PCR
product was identified by hybridization using 32P-labeled
sPARPcat as a probe, purified from agarose gel
(GenecleanTMSpin, Bio-101) and reamplified using AP2 and a
3' nested gene-specific primer containing SacI restriction
site (5'-CATCCCCACCAGCTCCTGCACTG-3') (5 cycles at 94 °C for 30 s, 72 °C for 3 min, 5 cycles at 94 °C for 30 s, 70 °C for
3 min; 25 cycles at 94 °C for 20 s, 68 °C for 3 min). The
PCR product (5'-sPARP) was then introduced in XbaI and
SacI restriction sites of pBluescript SK. The
total cDNA for sPARP-1 was sequenced according to the Sanger method
(42).
Radiation gene Hybrid Mapping--
The 5'-untranslated region of
the sPARP-1 gene was PCR amplified using the following pair of primers:
5'-GGTGAAGATGAGTAAGAAGATGG-3' (forward) and 5'-GGTGCTGTCAAGGAAGGAG-3'
(reverse). These primers yield a 235-bp product from mouse genomic DNA
but no amplification from hamster genomic DNA. The PCR screening of the
mouse T31 panel, which contained DNA from 100 radiation hybrid clones
plus two control DNAs (donor 129aa and recipient A23), was accomplished at Research Genetics, Inc., Huntsville, AL. To map the
sPARP-1 gene, the data vector, obtained by scoring the
radiation hybrid clones for a presence or absence of the 235-bp product
specific for the sPARP-1 gene, was submitted to the web
server of MIT Whitehead Institute.
Genomic Southern Blot Analysis--
Mouse genomic DNA was
isolated from fibroblast cells and subjected to Southern blotting
according to the standard procedure (43). Ten micrograms of the genomic
DNA was digested with each of EcoRI, HindIII,
BamHI, or PstI restriction endonucleases,
fractionated on a 0.7% agarose gel and alkaline-transferred to a
Hybond N+ membrane (Amersham Pharmacia Biotech). The blot
was hybridized with 32P-labeled 235-bp PCR product,
generated by the primers used for radiation hybrid mapping.
Activity-Western Blot Analysis--
Sample preparation was
carried out essentially as described earlier (44) except that protease
inhibitor mixture tablets (Complete, Mini; Roche Molecular Bichemicals,
Ontario, Canada) were used as antiproteases. The poly(ADP-ribosyl)ation
activity was visualized as described previously (45). Briefly, samples were resolved on a 10% SDS-polyacrylamide gel, the gel was then soaked
at 37 °C in Tris glycine running buffer containing 0.7 M
2-mercaptoethanol and the proteins were transferred onto a Hybond C
nitrocellulose membrane (Amersham Pharmacia Biotech). After refolding
of proteins in renaturation buffer (50 mM Tris/HCl, pH 8.0, 100 mM NaCl, 1 mM dithiothreitol, 0.3% Tween),
the enzymatic reaction was started by incubating the membrane in the
same buffer supplemented with 2 µg/ml DNase I-activated DNA, 2 mM Mg2+, 20 µM Zn2+,
and 100 µM NAD+. Noncovalently bound pADPr
was removed by four washes with SDS wash buffer (50 mM
Tris/HCl, pH 8.0, 100 mM NaCl, 1 mM
dithiothreitol, 2% SDS). Finally, covalently automodified proteins
were detected by using a 1:500 diluted monoclonal antibody 10H,
directed against the pADPr (46) and visualized with 1:5000 diluted
secondary peroxidase-coupled anti-mouse antibody (Jackson
ImmunoReasearch) and enhanced chemiluminiscence reagent (Renaissance
ECL-Plus, DuPont). The chemical and enzymatic stripping of the
activity-Western blots were performed as described earlier (44).
Development of Antibody and Western Blotting of sPARP-1--
A
peptide (MPSKEDAVEHFMKLY) chosen from the predicted amino acid sequence
of sPARP-1 was synthesized on a four-branch Multiple Antigenic Peptide
resin (47, 48) using a 433A Applied Biosystems Peptide Synthesizer with
Fast-MOC chemistry and injected into a rabbit to raise the polyclonal
antipeptide antibody (LR98-196) directed against sPARP-1. The
antiserum was affinity purified on Affi-Gel 10 resin (Bio-Rad) coupled
with the same peptide. For Western blot analysis of sPARP-1, nuclear
proteins from PARP-1 (+/+) and PARP-1 (/) cells were resolved on an
8% SDS-polyacrylamide gel, electrotransferred onto a nitrocellulose
membrane (Amersham Pharmacia Biotech), and blocked in
phosphate-buffered saline containing 5% skimmed milk and 0.1% Tween.
Blots were incubated overnight with affinity purified LR98-196 diluted
1:100 followed by goat anti-rabbit antibody coupled to peroxidase
(Jackson ImmunoReasearch) diluted 1:2500 in blocking solution.
Immunocomplexes were detected by ECL (Renaissance ECL-Plus, DuPont).
Cell Fractionation--
Cells grown to 80% confluence in 100-mm
dishes were incubated for 7 min with 500 µl of buffer A (300 mM sucrose, 10 mM Tris-HCl, pH 7.8, 3 mM MgCl2, 1 mM Na2EDTA,
2 mM 
-mercaptoethanol, CompleteTM Mini
antiprotease mixture (Roche Molecular Biochemicals)) supplemented with
0.03% Nonidet P-40. After homogenization with a glass-Teflon homogenizer (20 strokes), the suspension was centrifuged for 5 min at
800 × g. The cytoplasmic supernatant was removed and
the pelleted nuclei were washed three times in 500 µl of buffer A. The nuclear proteins were solubilized in SDS-polyacrylamide gel electrophoresis reducing loading buffer (49), sonicated, and kept at
80 °C for further analyses. All the steps were performed at
4 °C.
Purification of sPARP-1--
PARP-1 (/) cells were grown in
basal Dulbecco's modified Eagle's medium without methionine and
supplemented with 400 µCi of [35S]methionine (1,000 Ci/mmol; NENTM Life Science Products, Inc.). The nuclear
fraction prepared as described above was resuspended in column loading
buffer (50 mM Tris-HCl, pH 7.4; 1M NaCl, 10 mM
2-mercaptoethanol, one CompleteTM Mini protease inhibitor
mixture tablet (Roche Molecular Biochemicals) per 10 ml of buffer) and
sonicated on ice. The protein lysate was loaded on an Econo-Column
containing 400 µl of equilibrated 3-AB-Affi-Gel matrix prepared as
described elsewhere (50). The column was washed with 10 ml of
equilibration buffer followed by elution with a 10 mM
solution of 3-methoxybenzamide. Fractions of 200 µl were recovered
and 30 µl of each fraction was analyzed by autoradiography after a
10% SDS-polyacrylamide gel electrophoresis.
PADPr Blot Analysis--
PADPr was analyzed by electrophoresis
at 400 V on a 20% polyacrylamide gel essentially as described
elsewhere (51). PADPr was transferred onto a Hybond N+
membrane (Amersham Pharmacia Biotech) and immunodetected as described previously (52).
Immunofluorescence Microscopy--
PARP-1 (/) cells, grown
to confluence on coverslips, were treated with 200 µM
MNNG for 20 min, briefly washed with ice-cold phosphate-buffered
saline, and fixed in acetone/methanol (0.3/0.7, v/v) at 20 °C for
10 min. After saturation in phosphate-buffered saline containing 10%
non-fat milk and 0.1% Tween, cells were incubated in the same buffer
with the polyclonal antibody LP96-10 diluted 1:250 (52), and
subsequently with a 1:50 diluted secondary Texas Red-coupled
anti-rabbit antibody (Jackson ImmunoResearch). Slides were analyzed by
indirect fluorescence microscopy. For the immunoanalysis using the
polyclonal anti-peptide antibody LR98-196, the procedure was
essentially the same, except that PARP-1 (/) cells were not treated
with alkylating agent and the secondary antibody was fluorescein
isothiocyanate-conjugated anti-rabbit antibody.
DNA Breaks Measurement--
PARP-1 (/) cells grown to 80%
confluence in 6-well plates, were treated for different time intervals
with 100 µM MNNG as described above. The genomic DNA was
extracted and prepared as described previously (53). The 3' end of DNA
breaks was radiolabeled by a nucleotide exchange reaction catalyzed by
the T4 DNA polymerase (Amersham Pharmacia Biotech) in the presence of
[32P]dCTP (NENTM Life Science Products, Inc.)
and quantified according to the method of Legault et al.
(53).
| |
RESULTS |
PADPr Synthesis in PARP-1(/) Cells after Genotoxic
Treatment--
Treatment of cells with alkylating agents such as MNNG
usually leads to pADPr synthesis. In contrast, other genotoxic agents, like UVC for instance, do not significantly induce the synthesis of
pADPr. PARP-1 is the only well characterized enzyme responsible for the
synthesis of pADPr following DNA damage. To determine whether PARP-1 is
solely responsible for pADPr synthesis, we have studied this activity
in PARP-1 (/) cells. Wild type and mutant cells were treated with
100 µM MNNG or UVC irradiated with a dose of 30 J/m2 for different periods of time up to 60 min. After
purification by affinity chromatography, the amount of pADPr at each
time point was measured by an immunodot blot assay as described earlier
(39, 40). Table I shows the peak levels
of pADPr found in PARP-1 (/) and PARP-1 (+/+; F20) cells. Treatment
of cells for 20 min with 100 µM MNNG results in a
synthesis of 5.85 pmol of pADPr per mg of DNA for PARP-1 (/) cells,
whereas 60 pmol of pADPr per mg of DNA are synthesized by PARP-1 (+/+;
F20) cells. UVC treatment of PARP-1 (/) cells triggers the
synthesis of 12 pmol/mg of DNA; however, no increase of pADPr synthesis
was observed in PARP-1 (+/+; F20) cells. In all cases pADPr synthesis
was completely inhibited by the addition of 2 mM
1,5-dihydroxyquinolinediol, a PARP-1 inhibitor (54). In addition, our
results indicate that pADPr synthesis is not the only cellular process
responsible for NAD+ consumption since the measured
NAD+ depletion does not correlate directly with the amount
of synthesized pADPr. Thus, in accordance with Shieh et al.
(30), we observed pADPr production accompanied by NAD+
depletion in PARP-1 (/) cells suggesting the presence of an enzyme
bearing poly(ADP-ribose) polymerase activity.
View this table:
[in this window]
[in a new window]
|
Table I
Quantification of poly(ADP-ribose) and of NAD+ in PARP-1
(/) and PARP-1 (+/+) cells following genotoxic treatments
Cells were treated with either MNNG (100 µM) or UVC (30 J/m2). The amount of purified pADPr as well as purified
NAD+ was determined as described under "Experimental
Procedures." The data shown are the mean ± S.E. of at least
three independent experiments. All these results are expressed in
function of the amount of DNA. The amounts of synthesized pADPr are net
values obtained after subtraction of the basal level for the
corresponding cell line.
|
|
A Nuclear Protein of 60 kDa Is Responsible for pADPr Synthesis in
PARP-1 (/) and Is Also Present in PARP-1 (+/+)
Cells--
Activity-Western blots were performed with whole cell
extracts from PARP-1 (/) and PARP-1 (+/+; L) cells (Fig.
1A). To detect the pADPr
synthesized during the reaction, we used a monoclonal antibody directed
against pADPr (46). These activity-Western blots were conducted in the
presence or absence of NAD+. Purified bovine PARP-1 was
loaded in the left lane as a positive control. As expected, in the
presence of NAD+, the major band in PARP-1 (+/+; L) cells
appears at 113 kDa as a result of the catalytic activity of PARP-1.
However, it is accompanied by a second band corresponding to a protein
with a molecular mass of approximately 60 kDa that possesses
poly(ADP-ribose) polymerase activity (Fig. 1A, lanes 2, 4,
and 6). Only this latter protein is present in PARP-1
(/) cells (Fig. 1A, lanes 1, 3, and 5). The
signal intensity of 113- and 60-kDa bands depends on the amount of
loaded cells and therefore manifests that these two signals represent
specific enzymatic activities. In the absence of NAD+, no
enzymatic activities were detected, clearly demonstrating that, as for
PARP-1, NAD+ is a substrate for the 60-kDa enzyme.
Moreover, when these activity-Western blots were performed in the
presence of 1 mM 3-AB, a PARP-1 inhibitor, pADPr synthesis
was inhibited in purified bovine PARP-1 (a positive control for
inhibition), and in the samples isolated from PARP-1 (+/+; L) and
PARP-1 (/) cells (data not shown). Thus, the 60-kDa enzyme, like
PARP-1 appears to bind 3-AB leading to its inhibition.
View larger version (74K):
[in this window]
[in a new window]
|
Fig. 1.
Detection and localization of the enzyme
responsible for poly(ADP-ribose) formation in PARP-1 (/)
cells. A, presence of a 60-kDa protein bearing
poly(ADP-ribose) polymerase activity in both wild type and cells
lacking PARP-1. The cellular protein extracts equivalent to 400,000 (lanes 1 and 2), 300,000 (lanes 3 and
4), and 200,000 (lanes 5 and 6) cells
were resolved on a 10% polyacrylamide gel and electroblotted onto a
nitrocellulose membrane. After renaturation of proteins and
poly(ADP-ribose) synthesis in the presence of NAD+,
membranes were probed with a monoclonal antibody 10H directed against
pADPr. A protein with an apparent molecular mass of 60 kDa which
synthesize pADPr is present in both PARP-1 (/) (lanes 1, 3, and 5) and PARP-1 (+/+) (lanes 2, 4, and
6) cells. Purified bovine PARP-1 (0.1 units) was run as a
positive control (first lane). B, detection of
poly(ADP-ribose) polymerase activity in fractionated PARP-1 (/)
cells by activity-Western blot. The nuclear (left panel) and
cytoplasmic (right panel) fractions prepared from 200,000, 150,000, 75,000, and 50,000 PARP-1 (/) cells (lanes 1, 2, 3, and 4, respectively) were analyzed for the presence
of poly(ADP-ribose) polymerase activity by activity-Western blot. The
60-kDa protein able to form poly(ADP-ribose) is found in the nuclear
fraction. C, in situ immunodetection of
poly(ADP-ribose) synthesized in PARP-1 (/) cells by indirect
immunofluorescence microscopy. PARP-1 (/) cells grown under
standard conditions were mock-treated (panels 2 and
4) or treated (panels 1 and 3) with
200 µM MNNG for 20 min. Poly(ADP-ribose) was detected by
a polyclonal rabbit antibody followed by Texas Red-conjugated
anti-rabbit antibody. The slides were examined either by phase-contrast
(lanes 1 and 2) or fluorescence microscopy
(panels 3 and 4). D, purification of
sPARP-1 from PARP-1 (/) cells. PARP-1 (/) cells grown with
[35S]Met were fractionated into cytoplasmic and nuclear
fractions. The nuclear fraction was briefly sonicated in the
chromatography loading buffer at 4 °C and loaded onto a 3-AB-coupled
Affi-Gel matrix column. The elution of sPARP-1 was done with 10 mM 3-methoxybenzamide. Loading, washing, and elution
fractions (lanes 1-4) were subjected to SDS-polyacrylamide
gel electrophoresis, and then the gel was exposed for
autoradiography.
|
|
To determine the subcellular localization of the novel poly(ADP-ribose)
polymerase activity, PARP-1 (/) cells were fractionated into nuclei
and cytoplasm and activity-Western blots were performed on these two
fractions. The 60-kDa enzyme was found predominantly in the nuclear
fraction (Fig. 1B). The intensity of the band at approximately 60 kDa increases proportionally to the amount of loaded
proteins, indicating that the signal is due to a specific catalytic
activity. The faint signal in the cytoplasmic fraction may come from a
minor nuclear contamination. When similar activity-Western blots were
performed without NAD+, no band was observed (data not
shown). The subcellular localization of this activity was also studied
in situ using indirect immunofluorescence. PARP-1 (/)
cells were grown to confluence on coverslips, treated with 200 µM MNNG for 20 min, and stained with a polyclonal
antibody directed against pADPr and subsequently with a secondary,
Texas Red-conjugated antibody. No signal was observed in untreated
cells (Fig. 1C, panel 4), but following MNNG treatment,
pADPr was detected in the cell nuclei (Fig. 1C, panels 1 and
3). Thus, in agreement with the nuclear localization of the
60-kDa protein (Fig. 1B), these immunocytochemistry studies
indicate that the poly(ADP-ribose) polymerase activity is present
exclusively in the nucleus of PARP-1 (/) cells.
Based on our findings showing a nuclear localization and an inhibition
of its enzymatic activity by 3-AB, we carried out the purification of
the 60-kDa protein as described under "Experimental Procedures"
(Fig. 1D). Since the elution was made with a solution of
3-methoxybenzamide, another nicotinamide analogue and inhibitor of
poly(ADP-ribosyl)ation, the purified protein was not active as
confirmed by activity-Western blot (data not shown).
sPARP-1 cDNA and Chromosomal Localization of sPARP-1
Gene--
By using total RNA purified from PARP-1 (/) fibroblast
cells and primers chosen among the most highly conserved sequences of
PARP-1, a partial cDNA was isolated and then subcloned into pBlueScript SK. The rapid amplification of
cDNA ends technique was employed to isolate the full-length
cDNA. The 1732-bp complete cDNA was sequenced and found to be
identical to the catalytic domain of the mouse PARP-1. The complete
sequence is available in GenBank under the accession number AF126717.
An open reading frame in the DNA sequence lies between nucleotides 252 and 1730. The deduced amino acid sequence of this open reading frame
corresponds to the catalytic domain of the mouse PARP-1 and contains
all the amino acids required for the binding of NAD+ and
for the poly(ADP-ribose) polymerase activity. Therefore we called this
new isolated protein sPARP-1 for short
poly(ADP-ribose) polymerase 1. The nucleotide sequence from 1 to
251 corresponds to the 5'-untranslated region of the mRNA of
sPARP-1 and is different from the 5'-untranslated region sequence of
the mouse PARP-1. At position 141 in the 5'-untranslated region,
upstream of the predicted translation start point, an in-frame stop
codon was found indicating that the cloned cDNA contains the entire
open reading frame.
The similarity between the cDNA sequences of sPARP-1 and the
catalytic domain of PARP-1 suggested that only one gene encoded for
these two proteins. Therefore we carried out the chromosomal mapping of
the sPARP-1 gene by radiation hybrid analysis. A
comprehensive radiation hybrid map of the mouse genome is presently
available (55). The mouse T31 panel was screened in duplicate using PCR primers amplifying the 235-bp product from the 5'-untranslated region
sequence of sPARP-1 to identify the clones containing the sPARP-1
locus. The linkage analysis of the raw data mapped our marker on the
chromosome 1 of the Whitehead Institute (MIT) framework map between the
markers D1Mit459 and D1Mit509, 101 centimorgans (LOD>3.0). This locus
corresponds to the cytogenetic band 1H5 where is previously localized
the mouse PARP-1 gene and thereby indicates that sPARP-1
mRNA originates from this gene.
Concurrently to the RH mapping, we performed Southern blot analysis of
digested mouse genomic DNA as described under "Experimental Procedures." This analysis revealed the presence of one or at most
two bands consistent with a single copy gene as well (data not shown).
Identification of sPARP-1 as the Enzyme Responsible for the
Poly(ADP-ribose) Polymerase Activity Found in PARP-1 (/)
Cells--
Previous studies by others have shown that the 40-kDa
carboxyl-terminal end of PARP-1 contains the amino acids essential for its catalytic activity and is sufficient to support poly(ADP-ribose) polymerase activity (19). Since sPARP-1 virtually represented the
catalytic domain of PARP-1, it was expected to have poly(ADP-ribose) polymerase activity and, as confirmed by in vitro expression
of sPARP-1 cDNA (data not shown), sPARP-1 was competent to catalyze pADPr formation.
To determine whether sPARP-1 is identical to the 60-kDa protein
responsible for poly(ADP-ribosyl)ation seen in PARP-1 (/) and
PARP-1 (+/+) cells (Fig. 1, A and B), we
developed a polyclonal antibody (LR98-196) directed against a peptide
from the deduced amino acid sequence of the catalytic domain of
sPARP-1. It is noteworthy that the peptide chosen to develop the
LR98-196 antibody is absent in all of the recently described PARP
homologues. Western blot analysis using this antibody and
activity-Western blotting were performed in parallel, on the same
membrane, with nuclear protein extracts from PARP-1 (+/+) and PARP-1
(/) cells (Fig. 2A).
sPARP-1 is detected by the LR98-196 antibody in both PARP-1 (+/+) and
PARP-1 (/) nuclear extracts and accurately comigrates with the
protein bearing poly(ADP-ribose) polymerase activity. Furthermore, the
staining of methanol-fixed PARP-1 (/) cells with anti-sPARP-1
antibody revealed a stain pattern consistent with nuclear accumulation
of sPARP-1 (Fig. 2B). The immunodetection of sPARP-1 in the
nucleus either by Western blotting (Fig. 2A) or by
immunofluorescence (Fig. 2B) demonstrated that the
subcellular distribution of sPARP-1 coincided with the
poly(ADP-ribosyl)ation activity found in PARP-1 (/) cells (Fig. 1,
B and C). Taken together, our results demonstrate
that sPARP-1 is expressed in both PARP-1 (+/+) and PARP-1 (/) cells
and is in part responsible for the pADPr formation.
View larger version (76K):
[in this window]
[in a new window]
|
Fig. 2.
Identification of sPARP-1 as an active enzyme
responsible for poly(ADP-ribose) synthesis in PARP-1 (/)
cells. A, sPARP-1 is identical to the protein
responsible for poly(ADP-ribosyl)ation in PARP-1 (/) cells. Western
blot analysis (left panel) using anti-sPARP-1 antibody
(LP98-196), and activity-Western blot (right panel) with
nuclear extracts from PARP-1 (/) and PARP-1 (+/+) cells were
carried out in parallel, on the same membrane, in order to demonstrate
the identity of sPARP-1 with the protein catalyzing poly(ADP-ribose)
formation in PARP-1 (/) cells. B, sPARP-1 localizes in
the cell nucleus. PARP-1 (/) fibroblasts were fixed in
methanol-acetic acid and immunostained with anti-sPARP-1 antibody
followed by fluorescein isothiocyanate-conjugated anti-rabbit antibody.
Slides were examined either by phase-contrast (1) or
fluorescence microscopy (2).
|
|
The Activation of sPARP-1 Is DNA Break-independent--
As shown
in Table I, sPARP-1 is activated following treatment with DNA damaging
agents, suggesting that it could be activated by DNA strand breaks. To
test this hypothesis, activity-Western blots were performed with whole
PARP-1 (/) cell extract in the presence or absence of DNase
I-treated calf thymus DNA (Fig. 3, A and B, respectively). Equal amounts of purified
bovine PARP-1 were used as a positive control (Fig. 3, A and
B, left lane). As expected, the activity of purified PARP-1
increased in the presence of DNA treated with DNase I. In contrast, the
activity of sPARP-1 does not appear to be stimulated by activated DNA
since the intensity of the signal obtained with or without DNase
I-damaged DNA is similar (Fig. 3, A and B).
However, the signal intensity depends on the amount of cells loaded
showing that the activity which this signal represents is specific.
This result strongly suggests that, unlike PARP-1, sPARP-1 activation
is independent of DNA strand breaks.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 3.
DNA break-independent poly(ADP-ribose)
polymerase activity of sPARP-1. Activity-Western blot with PARP-1
(/) cells in the presence or the absence of DNase I-damaged DNA.
Protein extracts from PARP-1 (/) cells equivalent to 100,000 cells
(lanes 1), 200,000 cells (lanes 2), 300,000 cells
(lanes 3), and 400,000 cells (lanes 4) were
resolved on a 10% polyacrylamide gel. Purified bovine PARP-1 (0.1 units) was run as a positive control. The activity-Western blot was
performed in the presence (A) or absence (B) of 2 µg/ml DNase I-treated calf thymus DNA. C, time course of
DNA break generation and pADPr synthesis in PARP-1 (/) cells
following treatment with 100 µM MNNG. The amount of pADPr
per mg of DNA ( ) and the number of [32P]dCTP exchanged
per ng of DNA ( ) were determined for each of the indicated time
points of MNNG treatment. The data are representative of three
independent experiments.
|
|
To confirm this finding, PARP-1 (/) cells were treated with 100 µM MNNG during a time course. Kinetics of DNA strand
break formation and pADPr synthesis were determined and compared (Fig. 3C). Within the first 10 min of the MNNG treatment, the
number of DNA strand breaks increases four times without considerable synthesis of pADPr. However, while the increase in DNA strand breaks is
only 1.19-fold (from 10 to 20 min of the kinetic), pADPr synthesis
rises sharply to 5.85 pmol/mg of DNA, a more than 3-fold increase.
These kinetics thus indicate that the pADPr synthesis in PARP-1 (/)
cells does not correlate with the DNA strand break formation. Taken
together, these results demonstrate that DNA strand breaks do not
directly activate sPARP-1.
Analysis of pADPr Synthesized by sPARP-1--
PADPr from 1 × 108 cells treated with 200 µM MNNG for 20 min
was purified by affinity chromatography. After separation on a 20%
polyacrylamide gel, the pADPr was transferred onto a positively charged
nitrocellulose filter and immunodetected by a polyclonal antibody (Fig.
4, lane 1). Five pmol of pADPr
synthesized in vitro by purified PARP-1 was run as a
positive control (Fig. 4, lane 2). The electrophoresis was
performed until the bromphenol blue dye ran out of the gel. Under these
conditions, xylene cyanol dye migrates with the 20-mer pADPr (51). The
lengths of pADPr synthesized by sPARP-1 in PARP-1 (/) cells and
in vitro generated pADPr by purified PARP-1 appear to be
similar (Fig. 4). Moreover, purified poly(ADP-ribose) glycohydrolase
was able to hydrolyze pADPr synthesized by sPARP-1 suggesting
structural similarities between the pADPr synthesized by PARP-1 and by
sPARP-1 (data not shown). Our observations are consistent with a study
of Shieh et al. (30) which reports that pADPr synthesized by
PARP-1 (/) cells is indistinguishable from that of wild type cells
by several criteria. We therefore conclude that the characteristics of
pADPr synthesized by either of these poly(ADP-ribose) polymerases
cannot confer them a different cellular function.
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 4.
Immunodetection of poly(ADP-ribose)
synthesized in PARP-1 (/) cells. PADPr purified from 4 × 107 PARP-1 (/) cells (lane 1) and 5 pmol of
in vitro synthesized pADPr (lane 2) were
separated on a 20% polyacrylamide gel. After transfer onto a
nitrocellulose membrane, pADPr was immunodetected by polyclonal
antibody LP 96-10. XC, xylene cyanol.
|
|
| |
DISCUSSION |
This work was initiated to elucidate the presence of
poly(ADP-ribose) polymerase activity that we found in PARP-1 (/)
cells (Table I). We cloned from those cells a cDNA encoding a
protein that has an amino acid sequence identical to the catalytic
domain of the mouse PARP-1. Therefore we named this protein sPARP-1. sPARP-1 is expressed in both PARP-1 (/) and PARP-1 (+/+) cells. sPARP-1 gene belongs to the same locus as PARP-1
gene does. Therefore we suggest that PARP-1 and sPARP-1 are encoded by
the same gene. Multiple mRNAs encoded by a single gene have already
been reported for other mouse genes (56). There are three possible ways
for a gene to provide different mRNAs: an alternative splicing of the pre-mRNA, an alternative transcription starting point, or both.
PARP-1 (/) mice have been generated by inserting an in-frame neomycin cassette containing a stop codon and a synthetic
polyadenylation site between exon 2 and intron 2. Thus the full-length
PARP-1 mRNA is not present in PARP-1 (/) cells as shown
previously (25). In this case, an alternative splicing of PARP-1
pre-mRNA cannot occur in those cells. Therefore, we suggest that
sPARP-1 mRNA is a product of an alternative transcription
initiation site within the PARP-1 gene. The misdetection of
sPARP-1 mRNA in the earlier study by Wang et al. (25)
may result from the specificity of the labeled probe used for Northern
blot analysis or from its low transcription level.
Very recently, three novel PARP-1 homologues have been reported in
different organisms (34-36) thus forming a family of poly(ADP-ribose) polymerases. The common feature of these enzymes is a cluster of amino
acids called "PARP signature," which is implicated in the binding
and enzymatic processing of NAD+, their substrate, to form
pADPr. The poly(ADP-ribose) polymerases known so far can conditionally
be divided into two groups in respect to their mode of catalytic
activation: dependent and independent of DNA strand breaks. The
classical members PARP-1 and ZAP, isolated from mammals and maize,
respectively, and some of their truncated homologues, like NAP/APP (31,
32) from plants and mammalian PARP-2 (36), are known to require DNA
strand breaks for their activation. On the other hand, although the
mechanism of enzymatic activation of other family members as tankyrase
(33), vault PARP (57), and reported herein sPARP-1 is still ambiguous,
it does not appear to be mediated by DNA breaks. While sPARP-1
activation is DNA strand break-independent, we show that it is induced
following DNA damage. One interpretation of our findings may involve
the stimulation of sPARP-1 as a part of a biochemical and/or signaling pathway that operates in cells in response to genotoxic insults. In
this model, sPARP-1 enzymatic activation may result from
protein-protein interactions between up-stream acting components of the
pathway. Recently, PARP-1 enzyme activity was reported to be
up-regulated in 
-irradiated murine keratinocytes through the binding
of E1A adenovirus protein to the transcriptional coactivator p300/CBP, although this activation did not involve an increase in PARP-1 gene
expression (58). Alternatively, sPARP-1 induction following genotoxic
treatments may be regulated by post-translational modification of this
protein, like phosphorylation for instance. Indeed, PARP-1 activity was
previously shown to be modulated by phosphorylation during
Xenopus laevis oocyte maturation and the presence
of multiple sites for phosphorylation in PARP-1 has been suggested
(59). Whatever the mode of activation of sPARP-1 is, it would
apparently define the specificity of sPARP-1 action and function in
cells, since sPARP-1 actually represents the catalytic domain of PARP-1 and both enzymes synthesize identical pADPr.
The emergence of a family of PARP enzymes expands the field of
poly(ADP-ribosyl)ation and adds to the understanding of the pleiotropic
effects of PARP inhibitors observed in previous studies. Presently it
is difficult to determine which poly(ADP-ribose) polymerase is
responsible for the residual poly(ADP-ribosyl)ation observed in PARP-1
(/) cells. While it is now demonstrated that two additional
poly(ADP-ribose) polymerases, stimulated following genotoxic treatment,
are present in mammalian cells, sPARP-1 (this paper) and PARP-2 (36),
their particular contribution to the overall poly(ADP-ribosyl)ation
process is still unclear. Amé et al. (36) demonstrated
that DNase I-treated DNA enhances PARP-2 catalytic activity 15 times.
In contrast, we showed that sPARP-1 activity is not directly stimulated
by DNA strand breaks. In Fig. 3B, we present the kinetics of
pADPr synthesis and generation of DNA strand breaks in PARP (/)
cells treated with 100 µM MNNG. The results clearly show
that pADPr is synthesized in two steps. The first peak of pADPr
synthesis is concomitant with the generation of DNA strand breaks and
could therefore be attributed to PARP-2. In contrast, the second peak
of pADPr synthesis appears later, independently of the generation of
DNA lesions, and could be assigned to sPARP-1. We therefore suggest
that sPARP-1 may be implicated in the later stages of the cellular
response to DNA damage.
The physiological role of these novel poly(ADP-ribose) polymerases
remains to be elucidated. One can suggest that these enzymes could
serve as a backup for PARP-1, since PARP-1 (/) mice developed normally although poly(ADP-ribosyl)ation has been implicated in vital
processes such as replication, transcription, and differentiation. However, the hypersensitivity of PARP-1 (/) mice to genotoxic treatments suggests that these new enzymes cannot entirely compensate for PARP-1 functions. Despite the fact that the mode of activation of
DNA-independent PARP-1 homologues like sPARP-1, tankyrase and V-PARP
remain to be clarified, their presence and activity in cells reinforce
the importance of poly(ADP-ribosyl)ation. Generation of mice lacking a
single or more members of the PARP family will be a critical step
toward the elucidation of the particular function for these enzymes as
well as the role of poly(ADP-ribosyl)ation.
¶§,
, and
TOP
ABSTRACT
INTRODUCTION
These authors contributed equally to the results of this work.
